[Prev][Next][Index][Thread]
RE: Walls and Slots - was Re: APEC Rules
"Don't change the rules, change the environment"!
The first Portsmouth maze back in 1980 used corner pegs - and led to a
lot of grief.
The bright lads at ICL thought of the 'pins design' and that worked fine
for many years - until reliance was put on the 'freebie' maze from
Japan.
Let me remind you. Walls are held up by a pair of pins that slot into
(tiny) holes in the base. All 'posts' are virtual - incuded in the
wall. There are three lengths of wall, 'short' equal to the passage
width, 'medium' one 'virtual post' longer and 'long' that includes two
'virtual posts'. The only gaps/slots are where pairs of walls touch,
either butting together on a straight or at right angles on a corner.
Where a wall 'sticks out' there is no join at the end.
It's a lot cheaper to make, easier to transport and the maze base can be
bought, drilled and painted on site for a contest at the far end of an
air flight.
Cheers
John
-----Original Message-----
From: owner-micromouse@cs.rhul.ac.uk
[mailto:owner-micromouse@cs.rhul.ac.uk] On Behalf Of Derek Hall
Sent: Wednesday, 21 July 2004 8:00 AM
To: micromouse@cs.rhul.ac.uk
Subject: RE: Walls and Slots - was Re: APEC Rules
I couldn't agree more!
This in my view is what MicroMouse is all about, connecting the perfect
virtual world that exists inside a processor where you can accelerate to
the
speed of light in a picoseconds second, and the real world. That said I
think the slots should be mentioned in the rules, I agree most optical
systems can't see the indents. There have been problems with mechanical
sensors getting stuck in them and rendering a mouse that performs well
in
tests completely useless on competition day.
-----Original Message-----
From: owner-micromouse@cs.rhul.ac.uk
[mailto:owner-micromouse@cs.rhul.ac.uk]
On Behalf Of Peter Harrison
Sent: 20 July 2004 21:46
To: micromouse@cs.rhul.ac.uk
Subject: Walls and Slots - was Re: APEC Rules
David Otten wrote:
> "see". Since the post is the last thing the sensors see before an
open
> square, there is a chance that the slot confuses the mouse if the
> designer is not thinking about it.
And that is the key (so to speak) The plastic maze walls available in SE
Asia and now in the UK and elsewhere have these large slots in them.
Slots between walls and posts of up to 1mm or so [as in the rules] are
pretty well inevitable unless you can construct a maze to extremely good
tolerances. Actual slots may wel be bigger.
I worried for a while about making a really good maze until I realised
that it was pretty pointless. If I could not get my mouse to run in a
rough and ready maze, I would be designing it for an ideal, controlled
environment that may exist nowhere else in the world.
When testing, I sometimes just stand walls up deliberately out of place,
or at an angle or even place another wall inside the existing walls to
create a kind of chicane.
It is the designers responsibility to try and make the beast work under
the worst possible conditions. There is no fun or gain in complaining
because a mouse did less well due to the gaps being 0.25mm over size.
As they stand the rules make no mention of the slots, only gaps between
walls. But be honest - how many of you can detect the slots anyway? Tony
had trouble with resonant cavities for ultrasound. Dave mentioned a tiny
beamwidth for his sensors but still only shows the slots as a drop of a
few percent. My sensors throw an 8 degree beam which would be 15mm
across at the wall - I cant tell they are there. In any event, you
should have some idea where you are - compensate for when you expect to
be looking at a post or gap.
Take the issue of steps between parts of the maze floor. Fail to take
them properly into account and your mouse could be left with its wheels
in the air. My current mouse can handle a 1mm step [as in the rules] but
gets stranded on a 1.6mm step. That will have to be improved before I
run in a 'real' maze - just in case.
Pete Harrison
http://micromouse.cannock.ac.uk/